Barrios Et Al. (2012) International Taxation and Multinational Firm Location Decisions

8/16/2019 Barrios Et Al. (2012) International Taxation and Multinational Firm Location Decisions

1/13

International taxation and multinational rm location decisions☆

Salvador Barrios a, Harry Huizinga b,⁎, Luc Laeven c, Gaëtan Nicodème d

a European Commission, Joint Research Centre, IPTS b Tilburg University and CEPRc International Monetary Fund, Tilburg University and CEPRd European Commission, ULB, CEPR and CESifo

a b s t r a c ta r t i c l e i n f o

Article history:

Received 3 November 2008Received in revised form 18 June 2012

Accepted 18 June 2012

Available online 14 July 2012

JEL classi cation:

F23

G32

H25

R38

Keywords:

Corporate taxation

Withholding taxes

Dividends

Location decisions

Foreign direct investment

Multinationals

Using a large international rm-level data set, we examine the separate effects of host and additional parent

country taxation on the location decisions of multinational rms. Both types of taxation are estimated to have

a negative impact on the location of new foreign subsidiaries. The impact of parent country taxation is esti-

mated to be sizeable consistent with its international discriminatory nature. Our results show that interna-

tional double taxation by the parent country – despite the general possibility of deferral of taxation until

income repatriation – is instrumental in shaping the structure of multinational enterprise.

© 2012 Elsevier B.V. All rights reserved.

1. Introduction

With globalization and the progressive removal of barriers to trade,

an increasing number of companies develop international activities. To

access foreign markets, rms face a choice between producing goods at

home for exports and producing abroad. A host of tax and non-tax fac-

tors affect the decision whether to relocate production abroad. Among

the non-tax factors are the size of a foreign market, its growth pros-

pects, wage and productivity levels abroad, the foreign regulatory

and legal environment, and distance from the parent country (see

Görg and Greenaway (2004), Barrios et al. (2005), and Mayer and

Ottaviano (2007) for recent reviews). The impact of taxation on for-

eign direct investment (FDI) has been the subject of a sizeable liter-

ature, as reviewed by De Mooij and Ederveen (2006) and Devereux

and Maf ni (2007).

Studies of the effect of taxation on FDI location decisions generally

examine host country taxation to the exclusion of possible additional

parent country taxation. The contribution of this paper is to jointly con-

sider the impact of host and additional parent countrytaxation on mul-

tinational rm location decisions. As a rst level of taxation, the host

country may impose corporate income taxation on the income of localforeign subsidiaries. In addition, the host country could levy a non-

resident dividend withholding tax on the subsidiary's earnings at the

time they are repatriated to the parent rm. But taxation need not

stop at the host country level. The parent country can further choose

to levy a corporate income tax on the resident multinational's foreign

source income giving rise to internationaldouble taxation. We examine

the independent impactof all three levelsof taxation on the location de-

cisions of European multinationals over the period 1999–2003. Speci-

cally, we examine how these three levels of taxation affect in which

country a multinational headquartered in a certain country chooses to

locate a new foreign subsidiary.

Journal of Public Economics 96 (2012) 946–958

☆ The authorsthank JimHines andJoel Slemrod (the Editors), twoanonymous referees,

Wiji Arulampalam, Peter Finnigan, Andreas Hauer, Vanesa Hernandez Guerrero and

seminar participants at the Bancode España, the European Commission, the SolvayBrussels

School of Economics and Management, ZEW Mannheim, the CESifo area conference on

public sectoreconomics in 2009,and theCentrefor BusinessTaxationof OxfordUniversity

Summer Symposium in 2009 for valuable comments. The ndings, interpretations, and

conclusions expressed in this paper are entirely those of the authors. They should not be

attributed to the European Commission or the International Monetary Fund.

⁎ Corresponding author at: Department of Economics, Tilburg University, 5000 LE

Tilburg, Netherlands. Tel.: +31 13 4662623.

E-mail address: [email protected] (H. Huizinga).

0047-2727/$ – see front matter © 2012 Elsevier B.V. All rights reserved.

doi:10.1016/j.jpubeco.2012.06.004

Contents lists available at SciVerse ScienceDirect

Journal of Public Economics

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / j p u b e

http://dx.doi.org/10.1016/j.jpubeco.2012.06.004http://dx.doi.org/10.1016/j.jpubeco.2012.06.004http://dx.doi.org/10.1016/j.jpubeco.2012.06.004mailto:[email protected]://dx.doi.org/10.1016/j.jpubeco.2012.06.004http://www.sciencedirect.com/science/journal/00472727http://www.sciencedirect.com/science/journal/00472727http://dx.doi.org/10.1016/j.jpubeco.2012.06.004mailto:[email protected]://dx.doi.org/10.1016/j.jpubeco.2012.06.004


2/13

We have collected new detailed information on how parent country

tax systems interact bilaterally with corporate taxation and non-

resident withholding taxation in the host country. Specically, we use

information on whether or not countries tax the income of their multi-

nationalson a worldwide basis, and whether foreigntax credits arepro-

vided for non-residentwithholding taxes only or also for theunderlying

host country corporate tax (as, for instance, in the United States). As an

alternative to worldwide taxes, parent countries may partially or fullyexempt foreign source income from taxation. For example, Germany

exempts 95% of the foreign source income of German multinationals

from taxation.

Data on the structures of European multinationals, and in particular

on newforeign subsidiarylocations, areobtained from theAmadeus da-

tabase. This data set allows us to consider multinational companies

resident in a broad set of countries, each potentially having foreign sub-

sidiaries in many other countries. Thus, unlike earlier work this paper

considers multinational rm location choices in a multilateral setting

of N by N countries. In addition to being an innovative approach, this

multi-country framework is necessary to obtain suf cient variation in

additional parent country corporate taxation (not highly correlated

with host country corporate taxation) to be able to separately estimate

its impact on international location decisions.

Our results suggest that host country and additional parent coun-

try corporate income taxation both discourage the location of foreign

subsidiaries in a particular country. In our benchmark estimation, the

estimated negative impact of the two types of taxation – as derived

from statutory tax information – is about of equal size. Our nding

of a signicant role for the additional parent country tax in foreign

subsidiary taxation is new, and perhaps surprising. Many countries

allow parent country taxes on foreign source income to be deferreduntil dividend repatriation, reducing the scope for these taxes to af-

fect location decisions. All the same, we nd that foreign subsidiary

location – and the resulting international ownership pattern of sub-

sidiaries – is sensitive to additional parent country taxation. Multina-

tional rms act on an apparently signicant incentive to bring about

an international ownership pattern of subsidiaries that is internation-

ally tax-ef cient.

We perform several robustness checks to better understand the role

of host and parent country taxation in determining foreign subsidiary lo-

cation. We nd that location decisions of highly protable foreign sub-

sidiaries are less responsive to both host and parent country taxation

than location decisions of less protable foreign subsidiaries, perhaps be-

cause high protability reects location- or owner-specic rents that

would be destroyed by an alternative locationor nationalowner. Further,

Table 1

Corporate taxation and double tax relief methods for dividends received in European countries in 2003.

Corporate tax rate including local

taxes and surcharges (%)

Treatment of foreign dividends

from treaty countries

Treatment of foreign dividends

from non-treaty countries

Austria 24 Exemption Exemption

Belgium 33.99 Exemption (up to 95%) Exemption (up to 95%)

Bulgaria 23.5 Indirect credit Direct credit

Croatia 20 Exemption Exemption

Cyprus 15 Exemption Exemption

Czech Republic 31 Indirect credit DeductionDenmark 30 Exemption Exemption

Estonia 26 Indirect credit Indirect credit

Finland 29 Exemption Direct credit

France 35.43 Exemption (up to 95%) Exemption (up to 95%)

Germany 39.59 Exemption (up to 95%) Exemption (up to 95%)

Greece 35 Indirect credit Indirect credit

Hungary 19.64 Exemption Exemption

Iceland 18 Exemption Exemption

Ireland 12.5 Indirect credit Indirect credit

Italy 38.25 Exemption (up to 60%) Exemption (up to 60%)

Latvia 19 Exemption Exemption

Lithuania 15 Exemption Exemption

Luxembourg 30.38 Exemption Exemption

Malta 35 Indirect credit Indirect credit

Netherlands 34.5 Exemption Exemption

Norway 28 Indirect credit Indirect credit

Poland 27 Indirect credit Direct CreditPortugal 33 Direct credit Direct credit

Romania 25 Indirect credit Indirect credit

Russia 24 Direct credit No relief

Slovak Republic 25 Indirect credit No relief

Slovenia 25 Exemption Exemption

Spain 35 Exemption Indirect credit

Sweden 28 Exemption Exemption

Switzerland 21.74 Exemption Exemption

Turkey 33 Indirect credit Direct credit

United Kingdom 30 Indirect credit Indirect credit

Notes: Corporate tax rate denotes the statutory corporate tax rate including local taxes and surcharges. Statutory corporate tax rate in Estonia is 0% on

retained earnings but a distribution tax of 26% is applied on distributed pro t. Corporate tax rate of France includes a 3% social surcharge and a special

3.3% surcharge for large companies. Corporate tax rate of Germany includes a solidarity surcharge of 5.5%, an average deductible trade tax of 16.14%, and

an exceptional surcharge of 1.5%. Corporate tax rate of Hungary includes a deductible local business tax. Corporate tax rate of Ireland applies to trading

activities. For non-trading activities, the rate is 25%. Corporate tax rate of Luxembourg includes employment surcharges and local taxes. Corporate tax

rate of Switzerland applies to the canton of Zurich and includes cantonal and local taxes in Zurich. Treatment of foreign dividends refers to double tax

relief convention used by parent country. Foreign subsidiaries are assumed to be fully owned. The indirect credit system applies to foreign dividendsfrom treaty countries in Poland as long as the holding stake is at least 75% for the past 2 years and there exists a bilateral treaty or the EU

parent-subsidiary directive applies. In Portugal, foreign dividends from treaty countries are exempt from corporate taxes if the EU parent-subsidiary di-

rective applies, but the foreign withholding tax is not creditable. In some countries, dividend income is exempt from taxes up to a certain percentage,

indicated between brackets. Source: International Bureau of Fiscal Documentation.

947S. Barrios et al. / Journal of Public Economics 96 (2012) 946 –958


3/13

location choicesof foreign subsidiarieswithlowxed assetsare more re-

sponsiveto both host andparent country taxation. This could reect that

high xed assets are a barrier to choosing an alternative physical loca-

tion. Finally, we nd that international location decisions are relatively

sensitive to international double taxation – including the additional par-

ent country taxation– if theparent country does notallowthe deferralof

foreign source income until dividend repatriation.

Existing studies of the inuence of corporate taxes on multina-

tionals’ location have in general paid little attention to the role played

by parent country taxes. For instance, Devereux and Grif th (1998) in-

vestigate how host country taxationaffects the subsidiary location deci-

sions of US multinationals in several large European countries (France,

Germany, and the United Kingdom) over the period 1980–1994. Theynd that – conditional on the choice to locate production abroad –

host country average effective tax rates (but not marginal effective tax

rates) are important in determining foreign location choice, even if tax-

ation does not appear to affect the earlier choice to locate abroad or to

export. Buettner and Ruf (2007) in turn nd that location choices of

German multinationals across 18 potential host countries between

1996 and 2003 are affected more by host country statutory tax rates

thaneffective average tax rates, while they nd no effect of marginal ef-

fective tax rates.

Several authors, however, have previously found a role for addi-

tional parent country taxation to affect the location of FDI. For US

multinationals, Kemsley (1998) nds that the host country tax only

affects the ratio of US exports to foreign production over the period

1984–

1992 if the multinationals nd themselves in excess credit

positions.1 Analogously, a role of parent country taxation in affecting

FDI into the United States is found by Hines (1996) who shows that

foreign countries with worldwide taxation invest relatively much in

US states with high state taxes. This reects that multinationals locat-

ed in countrieswith worldwide taxation may be able to obtainforeign

tax credits for US state corporate income taxes. Altshuler and Grubert

(2001) evaluate the implications of a hypothetical switch to dividend

exemption by the US to conclude, in contrast, that there is no evi-

dence that such a change would materially affect international loca-

tion decisions of US multinationals. Egger et al. (2009) construct an

effective tax rate on a bilateral basis that reects overall host and

home country taxation, and nd that this bilateral effective tax rate

has a negative impact on bilateral FDI stocks after controlling forhost and parent country unilateral effective tax rates.2 However,

none of these papers studies a rm's location decision in a multilater-

al setting of N by N countries, as we do.

Table 2

Bilateral dividend withholding tax rates in Europe in 2003.

Sub sidiary c ountr y Par ent c ou ntr y A ustr ia Belgium B ulga ria C roat ia C yp rus C zech R ep . Denmar k Estonia Finla nd France Germa ny

Austria X 0 0 0 10 10 0 5 0 0 0

Belgium 0 X 10 10 10 5 0 25 0 0 0

Bulgaria 0 10 X 5 5 10 5 15 10 5 15

Croatia 0 10 5 X 10 5 5 15 5 5 0

Cyprus 0 0 0 0 X 0 0 0 0 0 0

Czech Rep. 10 5 10 5 10 X 15 5 5 10 5

Denmark 0 0 5 5 10 15 X 5 0 0 0Estonia 0 0 0 0 0 0 0 X 0 0 0

Finland 0 0 10 5 29 0 0 0 X 0 0

France 0 0 5 5 10 10 0 5 0 X 0

Germany 0 0 15 15 10 5 0 5 0 0 X

Greece 0 0 0 0 0 0 0 0 0 0 0

Hungary 10 10 10 10 5 5 5 20 5 5 5

Iceland 15 15 15 15 15 5 0 5 0 5 5

Ireland 0 0 0 20 0 0 0 0 0 0 0

Italy 0 0 10 10 15 15 0 5 0 0 0

Latvia 10 10 10 5 10 5 5 5 5 5 5

Lithuania 15 15 15 5 15 5 5 0 5 5 5

Luxembourg 0 0 0 20 20 0 0 20 0 0 0

Malta 0 0 0 0 0 0 0 0 0 0 0

Netherlands 0 0 5 0 25 0 0 5 0 0 0

Norway 5 15 15 15 0 5 0 5 0 0 0

Poland 10 10 10 5 10 5 5 5 5 5 5

Portugal 0 0 15 30 30 15 0 30 0 0 0Romania 10 5 10 5 10 10 10 10 5 10 10

Russia 5 15 15 5 0 10 10 15 5 5 5

Slovak Rep. 10 5 10 5 10 5 15 15 5 10 5

Slovenia 5 5 15 15 10 5 5 15 5 5 15

Spain 0 0 5 15 15 5 0 15 0 0 0

Sweden 0 0 0 0 0 0 0 0 0 0 0

Switzerland 5 10 5 35 35 5 0 35 5 5 5

Turkey 16.5 16.5 16.5 11 16.5 16.5 16.5 16.5 16.5 16.5 16.5

United Kingdom 0 0 0 0 0 0 0 0 0 0 0

Notes: Withholding tax rates apply to dividends paid by fully owned subsidiaries in subsidiary country to parent rm. Bilateral tax treaties are taken into account. Parent-Subsidiary

Directive is binding between EU Member States and provides exemption from withholding tax if equity holding is at least 25%. The reported gures assume an equity holding in the

subsidiary of at least 25%. In Ireland, subsidiaries owned by parent companies resident in EU or treaty countries are exempt from withholding tax provided that they are not under

the control of persons not resident in such countries. In Italy, authorities can provide a refund equal to the tax claimed limited to 4/9 of the Italian withholding tax if the recipient

can prove a tax is paid in his country on the dividend. In Luxembourg there is an exemption from withholding tax for EU and treaty partners if holding in company resident in

Luxembourg is at least 10%. Source: International Bureau of Fiscal Documentation.

1 US multinationals are subject to worldwide taxation in the United States. Thus,

they have to pay tax in the United States on their foreign-source income, subject to

the provision of a foreign tax credit for taxes already paid in the host country. The for-

eign tax credit, in practice, is limited to the amount of US tax due on the foreign-source

income. This implies that the overall tax on the foreign income is the host country tax if

this tax exceeds the US tax, while it is the US tax if this tax is the higher of the two. US

taxes on foreign source income can be deferred until the income is repatriated.2 Repatriation taxes more broadly can affect multinational rms' behavior. Desai et

al. (2001) analyze the effect of repatriation taxes on dividend payments by the foreign

af liates of US multinational rms to nd that 1% lower repatriation tax rates are asso-

ciated with 1% higher dividends.

948 S. Barrios et al. / Journal of Public Economics 96 (2012) 946 –958

http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-


4/13

Using information on international M&As, Huizinga and Voget

(2009) nd that the international ownership pattern that emerges re-

ects an effort to avoid international double taxation of foreignsource income. Desai and Hines (2002) argue that international dou-

ble taxation imposed by the US can explain inversions of existing US

multinationals, whereby the original US parent becomes a subsidiary

and the earlier foreign subsidiary becomes the new parent rm. Voget

(2011) nds that a multinational is more likely to relocate its head-

quarters abroad, if it is located in a parent country with high taxation

of foreign source income. This varied evidence on how international

double taxation affects the structure of the multinational rm is

consistent with a pattern of foreign subsidiary location that avoids

additional parent country taxation, as found in the present paper.

Huizinga et al. (2012) demonstrate that the international double

taxes that are triggered by an international M&A are to a large extent

capitalized into a lower takeover price. This also suggests that inter-

national double taxes are economically burdensome to multinationalrms, providing them with an incentive to own foreign subsidiaries

that are subject to low parent country taxation.

At a theoretical level, Hartman (1985) makes a distinction between

‘mature’ foreign af liates that are nanced at the margin by retained

earnings, and ‘immature’ foreign af liates that rely on funding from

their parent rms. The af liate's required rate of return is shown to re-

ect parent country taxation only if it is immature. Parent country tax-

ation thus should be particularly important for the newly established

foreign subsidiaries that are examined in this paper.

In the remainder of this paper, Section 2 describes the tax treat-

ment of the foreign source income of multinational rms. Section 3

discusses our rm-level data. Section 4 presents estimates of the im-

pact of international taxation on the location of foreign subsidiaries.

Finally, Section 5 concludes.

2. The international tax system

This section describes the corporate tax system applicable to amultinational company with foreign subsidiaries.3 Consider a multi-

national company with a parent located in home country p and a sub-

sidiary located in host country s. Both home and host countries may

tax the subsidiary's income. First, the host country may levy a corpo-

rate income tax at a rate t s on this income. Table 1 shows the statutory

corporate income tax rates for the 33 European countries in our sam-

ple for the year 2003.4 These statutory tax rates are those on distrib-

uted prots and include local taxes and applicable surcharges. In our

sample, the corporate tax rate for 2003 ranges from a low of 12.5% in

Ireland to a high of 39.6% in Germany.

Next, the host country levies a non-resident dividend withholding

tax at a rate ws on the subsidiary's net of corporate tax income upon

repatriation of this income to the parent. Table 2 provides informa-

tion on the applicable withholding tax rates on dividends paid byfully owned subsidiaries to their non-resident parents in 2003. For

example, a dividend paid by a Belgian subsidiary to its parent compa-

ny located in Estonia will be subject to a withholding tax of 25%, while

the withholding tax on a dividend paid by an Estonian subsidiary to

its Belgian parent company has a zero rate. The withholding tax

Greece Hungary Iceland Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Norway Poland

0 10 25 0 0 25 25 0 15 0 5 10

0 10 25 0 0 25 25 0 15 0 5 10

10 10 15 5 10 15 15 5 0 5 15 10

5 5 15 15 10 5 15 15 5 5 15 5

0 0 0 0 0 0 0 0 0 0 0 0

15 5 5 5 15 5 5 5 5 0 5 5

0 5 0 0 0 5 5 0 0 0 0 00 0 0 0 0 0 0 0 0 0 0 0

0 5 0 0 0 0 0 0 0 0 0 0

0 5 5 0 0 5 5 0 5 0 0 5

0 5 5 0 0 5 5 0 5 0 0 5

X 0 0 0 0 0 0 0 0 0 0 0

10 X 20 5 10 20 20 5 5 5 10 10

15 15 X 15 15 5 5 5 15 0 0 5

20 0 20 X 0 0 0 0 20 0 0 0

0 10 27 0 0 27 5 0 15 0 15 10

10 10 5 5 10 X 0 10 5 5 5 5

15 15 5 5 5 0 X 15 15 5 5 5

0 0 0 0 0 20 20 X 0 0 0 0

0 0 0 0 0 0 0 0 X 0 0 0

0 5 0 0 0 5 5 0 5 X 0 0

20 10 0 0 15 5 5 5 15 0 X 5

15 10 5 0 10 5 5 5 5 0 5 X

0 15 15 0 0 30 30 0 15 0 15 1510 5 10 3 10 10 10 5 5 5 10 5

15 10 15 10 5 15 15 10 15 5 10 10

15 5 15 0 15 10 10 5 5 0 5 5

15 10 15 5 10 5 5 5 15 5 15 5

0 5 5 0 0 15 15 0 15 0 10 5

0 0 0 0 0 0 0 0 0 0 0 0

5 10 5 10 15 5 5 0 35 0 5 5

16.5 11 16.5 16.5 16.5 16.5 11 16.5 16.5 16.5 16.5 11

0 0 0 0 0 0 0 0 0 0 0 0

3 See Huizinga et al. (2008) for a more detailed description of corporate tax systems

as they apply to multinational companies. This study is limited to the impact of the in-

ternational taxation of dividends on location decisions, even though the taxation of

other forms of income such as royalties or interest could also be important in corporate

location choices.4 For illustrative purposes, the tables report taxation data for the year 2003 only, al-

though we have collected these data for the entire period 1999–

2003.


http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-


5/13

rates for transactions involving two EU Member States currently are

zero on account of the EU Parent-Subsidiary Directive.5

The net of withholding tax dividend received by the parent com-

pany is in principle taxed in the parent country–subject to some

form of double tax relief as recommended by the OECD Model TaxTreaty or as prescribed by the EU Parent-Subsidiary Directive. Some

countries operate an exemption system. In this instance, the dividend

is not taxed in the parent country, if the provided exemption is full.

The overall international rate of taxation on the subsidiary's income

is then given by 1−(1− t s)(1−ws) or t s+ ws− t sws.

In other instances, the home country may tax the worldwide in-

come of its multinationals and subject the received dividend to corpo-

rate income taxation at a rate t p. Generally, a foreign tax credit is

provided for taxes paid in the host country, usually limited to the

amount of the home tax due on the foreign source income. Some

countries apply an indirect tax credit system under which both the

corporate tax and the withholding tax paid in the host country are

credited against the home corporate income tax. In case the home

country's corporate income tax t p is higher than the overall hostcountry tax rate t s+ ws−t sws, the rm pays income tax in the

home country at a rate t p− [t s+ws− t sws] so that the combined, ef-

fective tax rate is equal to t p. If instead the home country's corporate

income tax rate is lower than the overall host country's rate, the rm

is said to be in excess foreign tax credit and it will pay no further tax

in the home country (having reduced its home tax liability to zero by

using foreign tax credits). In this instance, the combined, effective tax

rate is t s+ws− t sws. In summary, for home countries with an indirect

tax credit system, the combined, effective tax rate is equal to max

[t p, t s+ ws− t sws].

Home countries may restrict the foreign tax credit to cover only

host country non-resident withholding taxes giving rise to a direct

tax credit system. In this case, the multinational has to pay tax in

the home country to the extent that t p exceeds ws and the combined,

effective tax rate is given by t s+(1−

t s) max[t p, ws].Alternatively, some home countries offer neither exemption nor a

foreign tax credit for taxes paid abroad, but instead allow foreign

taxes to be deducted from home country taxable corporate income.

This amounts to the deduction system with a combined, effective

tax rate of 1−(1− t s)(1−ws)(1−t p).

Finally, in some rather exceptional cases no double tax relief is

provided at all. With full double taxation, the combined, effective

tax rate becomes t s+ ws− t sws+t p.

Columns 3 and 4 of Table 1 indicate which doubletax relief system

is applied by European countries in the sample. As seen in the table,

some countries provide different double tax relief to treaty partners

and non-treaty countries. Thus, we need to know whether there

exist doubletax treaties among thecountries in oursample. On a bilat-

eral basis, this information is provided in Table 3 with the value 1 in-dicating the existence of such a treaty and 0 its absence. The table

indicates that for many countries the treaty network is not complete.

Forexample,in 2003 theCzechRepublic hasa treaty with allcountries

in the sample except Malta and Turkey. From Table 1, we see that this

implies that dividends from all foreign subsidiaries paid to a Czech

parent benet from an indirect tax credit, except for those paid by a

Maltese or a Turkish subsidiary where the deduction system applies.

Information from Tables 1–3 allows us to calculate the combined

effective tax rate on foreign dividends for any pair of home and host

countries. To x ideas, consider the case of a dividend paid by a Mal-

tese subsidiary to its Czech parent in 2003. Table 1 shows that the

statutory corporate tax rate in Malta is 35%. We infer from Table 2

that net prots paid as a dividend to a foreign company are never

subject to a non-resident dividend withholding tax in Malta. As

Subsidiary country Parent country Portugal Romania Russia Slovak Rep. Slovenia Spain Sweden Switzerland Turkey United Kingdom

Austria 0 15 5 10 5 0 0 0 25 0

Belgium 0 5 10 5 5 0 0 10 15 0

Bulgaria 10 10 15 10 15 5 10 5 10 10

Croatia 15 5 5 5 15 5 15 5 10 5

Cyprus 0 0 0 0 0 0 0 0 0 0

Czech Rep. 10 10 10 5 5 5 0 5 15 5

Denmark 0 10 10 15 5 0 0 0 15 0

Estonia 0 0 0 0 0 0 0 0 0 0

Finland 0 0 0 0 5 0 0 0 15 0

France 0 10 10 10 5 0 0 5 15 0

Germany 0 10 5 5 15 0 0 0 15 0

Greece 0 0 0 0 0 0 0 0 0 0

Hungary 10 5 10 5 10 5 5 10 10 5

Iceland 10 15 15 15 15 5 0 5 15 5

Ireland 0 0 0 0 0 0 0 0 20 0

Italy 0 10 5 15 10 0 0 15 15 0

Latvia 10 10 10 10 5 10 5 5 10 5

Lithuania 15 10 15 10 5 5 5 5 10 5

Luxembourg 0 0 0 0 0 0 0 0 20 0

Malta 0 0 0 0 0 0 0 0 0 0

Netherlands 0 0 5 0 5 0 0 0 5 0

Norway 10 10 10 5 15 10 0 5 20 5

Poland 10 5 10 5 5 5 5 5 10 5

Portugal X 15 15 30 30 0 0 15 30 0

Romania 10 X 10 10 5 10 10 10 10 10

Russia 10 15 X 10 10 5 5 5 10 10Slovak Rep. 15 10 10 X 5 5 0 5 5 5

Slovenia 15 15 10 5 X 5 5 5 15 5

Spain 0 5 5 5 5 X 0 10 15 0

Sweden 0 0 0 0 0 0 X 0 0 0

Switzerland 10 10 5 5 15 10 0 X 35 5

Turkey 16.5 16.5 11 5.5 16.5 16.5 16.5 16.5 X 16.5

United Kingdom 0 0 0 0 0 0 0 0 0 X

Table 2 (continued)

5 Note that in 2003 prior to their accession, many new EU Member States still

maintained non-zero rates vis-à-vis EU countries and vice versa.


http://-/?-http://-/?-http://-/?-http://-/?-


6/13

already mentioned, Table 3 indicates that no tax treaty was in force

on income repatriated from Malta to the Czech Republic in 2003 so

that from Table 1 we see that incoming foreign dividends benet

from a deduction system in Czech Republic. Finally, the same table

indicates that the applicable corporate tax rate in this country is

31%. From the formula above, the combined, effective tax rate equals

1−(1−0.35)×(1−0)×(1−0.31)=55.2%. This rate is considerably

higher than the Maltese corporate tax rate of 35%. This suggests

that the additional taxation of multinational

rms, in the form of withholding taxes and home country corporate income taxation, po-

tentially has an independent and signicant impact on international

location decisions.

Below, we will investigate the independent inuences of host

country corporate income and dividend withholding taxation and

home country corporate income taxation on corporate location deci-

sions. To make parameter estimates comparable across tax measures,

it is useful to construct all three tax measures as shares of the foreign

subsidiary's pre-tax income. The host country tax rate is already de-

ned as a share of the subsidiary's pre-tax income. Our withholding

tax measure will be (1− t s) ws to reect that the withholding tax ap-

plies to the subsidiary's income net of the host country corporate tax.

Finally, the additional parent country corporate income tax – as a

share of the subsidiary's pre-tax income – is computed as the differ-

ence between the combined, effective tax rate and t s+ ws− t sws.

We will dene the international tax to be the sum of the withholding

tax and additional parent country corporate tax both expressed as

shares of the subsidiary's pre-tax income. Equivalently, the interna-

tional tax is the difference between the combined, effective tax and

the host country corporate income tax.

Unlike host country corporateincome taxes, withholding taxesand

home country corporate income taxes are generally deferred until the

foreign source income is repatriated to the parent in the form of divi-

dends. Deferral reduces the present value of taxation. Thus, withhold-

ing taxes and home country corporate income taxes are expected to‘bite’ less than host country corporate income taxes. Whether the de-

ferral of withholding taxes and home country corporate income taxes

serves to make these taxes immaterial for location decisions is an em-

pirical matter. This is what we turn to in the empirical section below.

3. Multinational enterprise data

Data on the structure of multinational rms in Europe are taken

from the Amadeus database. This database provides standard ac-

counting data and data on ownership relationships within corporate

groups.6 We have data on multinational rms operating in 33 Europe-

an countries over the years 1999–2003. The ownership data enable us

to match European rms with their domestic subsidiaries and foreign

subsidiaries located in other European countries.7 We have owner-

ship information for the years 1999, 2001, and 2003. A rm is called

a subsidiary if at least 50% of the shares are owned by a single other

rm. A subsidiary is taken to be new in its year of incorporation.

In our benchmark analysis below, we consider 909 new foreign sub-

sidiaries. Information on the number of parent and subsidiary countries

involved in these new locations is provided in Panel A of Table 4. Our

benchmark sample excludes new locations where the parent company

becomes an intermediate company as it is a subsidiary itself of another

parent company. The United Kingdom with 115 new parent companies

has most new parent companies, followed by France with 84 new par-

ent companies. Eachsubsidiary has a home country (where its parent is

located) and a host country (whereit is located itself). For each country,the table lists the number of subsidiaries by home country and by host

country. The table indicates that, for example, France, the Netherlands,

and the United Kingdom are the home country to relatively many

subsidiaries. Hence,there are relatively manysubsidiaries with a parent

rm in one of these countries. Denmark, Germany and the United

Kingdom, on the other hand, are the host country to relatively many

subsidiaries.

Our subsequent empirical work on foreign subsidiary location

aims to predict the location of a new foreign subsidiary in 1 of 32 for-

eign European countries. The dependent variable, called subsidiary

location, takes on a value of one if a particular country is selected as

a subsidiary's location and it is zero otherwise.

Summary statistics on the subsidiary location variable, the tax vari-

ables, and some controls are provided in Panel B of Table 4 (see Table

A1 in the Appendix for variable denitions and data sources). The

26,648 observations reported in the table are identical to the number of

observations in the basic regression 1 of Table 5.8 The mean value of the

overall effective tax is 0.35. This mean effective tax, in effect, is the sum

of a mean host country tax of 0.30 and a mean international tax of 0.05.

Among the control variables, GDP bilateralis the ratio ofthe GDP of a

potential host country andthe sumof theGDPs of allother potential for-

eign (but not domestic) locations. This variable captures market size,

and it is expected to exert a positive impact on the probability of subsid-

iary location in a host country.

Contiguity is a dummy variable signaling a common border be-

tween host and home countries. A common border is expected to

make location in the host country more likely.

Difference in labor costs is the log of the ratio of labor costs in the

home country and labor costs in the host country, expressed in a com-mon currency. The impact of higher labor costs in the host country on

the probability of location is in principle ambiguous (see Kimino et al.

(2007) for a review of the empirical evidence). Higher labor costs in

the host country (or a lower difference in labor costs variable) are

expected to discourage location for a given level of labor productivity

in the host country. In contrast, they potentially encourage location to

the extent that they signal high labor skills and productivity that are

sought after by the multinational rm.

Economic freedom is an index of the extent of soundness of the legal

system, absence of trade barriers, absence of price controls, and transfers

and subsidies as a share of GDP. Economic freedom should make a coun-

try attractive as a subsidiary location.

Finally, EU membership is a dummy variable agging EU member-

ship of a prospective host country. EU membership, to the extent thatis signals commitment to high EU standards of dealing with foreign in-

vestors, could engender subsidiary location. Moreover, establishing a

subsidiary in one EU country allows non-EU rms to trade freely across

all other EU countries by taking advantage of the EU common market.

Panel C of Table 4 provides correlation coef cients among the lo-

cation, tax and control variables. Interestingly, location is positively

and signicantly related to the host country tax, but negatively and

signicantly to the international tax. The rst correlation possibly re-

ects that subsidiaries tend to be located in larger countries, which

tend to have relatively high corporate income taxes. In the table,

the host country tax is indeed positively and signicantly correlated

6 The database is created by collecting standardized data received from 50 vendors

across Europe. The local source for these data is generally the of ce of the Registrar

of Companies.7 The Amadeus database only contains information on European rms and we there-

fore only cover the European operations of the multinationals in our sample. In 2001,

intra-EU25 FDI ows amounted to 54% of all outward FDI ows from EU25 countries,

while intra-EU25 FDI stocks were 57% of outward FDI stocks of EU25 countries

(according to Eurostat). Thus, European multinationals have most of their FDI out-

standing in other European countries. All the same, our conditioning on a European lo-

cation of new foreign subsidiaries potentially affects estimated sensitivities of location

to taxation. In particular, by conditioning on European locations rather than locations

worldwide, we increase the expected probability of location in any one country and

potentially also the sensitivity of location to taxation. New foreign subsidiaries tend

to be large relative to the overall assets and capital of the multinational. In our sample,

a new foreign subsidiary on average represents 23% of the assets and 24% of the capital

of the expanded multinational rm including the new foreign subsidiary.

8 Note that the mean value of the location variable is not exactly 1/32 due to the ab-

sence of data for some specic combinations of countries and years.


http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-


7/13

with the GDP bilateral variable as an index of host country relative

size. The negative correlation between location and the internationaltax could reect that subsidiary location is chosen so as to mitigate in-

ternational double taxation. The host country tax and the internation-

al tax are negatively correlated, perhaps reecting the operation of

the foreign tax credit mechanism.

4. Empirical results on taxation and subsidiary location

In this section, we estimate the impact of host and additional parent

country taxation on the choiceof location of a newforeign subsidiary. In

our estimation, we explain subsidiary location choice only in the initial

year of establishment to ensure that observations regarding different

location episodes are independent. We condition on a foreign subsidi-

ary location, and hence examine the impact of international taxation

on the location choice among competing foreign locations.Our sample consists of new foreign subsidiaries that incorporate

in one of the years from 1999 to 2003. The data set encompasses 33

European countries, which implies that for a multinational resident

in a particular home country there are 32 foreign location options. Ac-

cordingly, for each new foreign subsidiary we construct 32 binary

variables that take on a value of one if the actual location is in a cer-

tain foreign country and zero otherwise. Regarding each potential lo-

cation for each new foreign subsidiary, there thus is a binary choice.

Location choice is assumed to be determined by the various countries’

tax rates and a range of other location or country characteristics.9 The

underlying binary choice model is estimated using the conditional

logit approach of McFadden (1974, 1976).

10

In this approach, explana-tory variables vary across alternative outcomes. All regressions include

year xed effects, which are not reported.11 We also control for possible

clustering of the errors at the level of the potentialhostcountry to allow

for the possibility that shocks to the attractiveness of host countries af-

fect more than a single subsidiary location choice.12 Specically, we re-

port heteroscedasticity-consistent standard errors that are adjusted for

clustering at the level of the host country.

In regression 1 of Table 5, we relate foreign subsidiary location only

to the effective tax rate. The number of observations is 26,648 rather

than 29,088 (or 909 new subsidiaries times 32 potential location coun-

tries) due to missing observations of the effective tax variable for some

combinations of countries and years. The effective tax variable enters

with a negative coef cient of −0.87 that is statistically insignicant.

Table 3

Existence of bilateral tax treaties for European country pairs in 2003.

Income From: Austria Belgium Bulgaria Croatia Cyprus Czech

Rep.

Denmark Estonia Finland France Germany Greece Hungary Iceland Ireland Italy

To:

Austria X 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1

Belgium 1 X 1 1 1 1 1 0 1 1 1 1 1 0 1 1

Bulgaria 1 1 X 1 1 1 1 0 1 1 1 1 1 0 1 1

Croatia 1 1 0 X 1 1 1 0 1 1 1 1 1 0 0 1

Cyprus 1 1 1 1 X 1 1 0 0 1 1 1 1 0 1 1

Czech Rep. 1 1 1 1 1 X 1 1 1 1 1 1 1 1 1 1

Denmark 1 1 1 1 1 1 X 1 1 1 1 1 1 1 1 1

Estonia 0 0 0 0 0 1 1 X 1 1 1 0 0 1 1 1

Finland 1 1 1 1 0 1 1 1 X 1 1 1 1 1 1 1

France 1 1 1 1 1 1 1 1 1 X 1 1 1 1 1 1

Germany 1 1 1 1 1 1 1 1 1 1 X 1 1 1 1 1

Greece 1 1 1 1 1 1 1 0 1 1 1 X 1 0 1 1

Hungary 1 1 1 1 1 1 1 0 1 1 1 1 X 0 1 1

Iceland 0 0 0 0 0 1 1 1 1 1 1 0 0 X 0 0

Ireland 1 1 1 0 1 1 1 1 1 1 1 0 1 0 X 1

Italy 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 X

Latvia 0 0 0 1 0 1 1 1 1 1 1 0 0 1 1 0

Lithuania 0 0 0 1 0 1 1 1 1 1 1 0 0 1 1 1

Luxembourg 1 1 1 0 0 1 1 0 1 1 1 1 1 1 1 1

Malta 1 1 1 1 1 1 1 1 1 1 1 0 1 0 0 1

Netherlands 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1

Norway 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

Poland 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1Portugal 1 1 1 0 0 1 1 0 1 1 1 1 1 1 1 1

Romania 1 1 1 1 1 1 1 0 1 1 1 1 1 0 1 1

Russia 1 1 1 1 1 1 1 0 1 1 1 0 1 0 1 1

Slovak Rep. 1 1 1 1 1 1 1 0 1 1 1 1 1 0 1 1

Slovenia 1 1 0 0 1 1 1 0 1 1 1 0 1 0 1 1

Spain 1 1 1 0 0 1 1 0 1 1 1 1 1 1 1 1

Sweden 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

Switzerland 1 1 1 0 0 1 1 0 1 1 1 1 1 1 1 1

Turkey 1 1 1 1 0 0 1 0 1 1 1 0 1 0 0 1

United Kingdom 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

Notes: This table addresses whether a tax treaty was in effect to deal with income received by countries listed in the rows and originating from countries listed in the columns.

Specically, 1 denotes that a bilateral tax treaty was applicable and 0 denotes that a tax treaty was not applicable. The table is not exactly symmetric as the dates of rst application

of a treaty may slightly differ between two treaty partners. Source: International Bureau of Fiscal Documentation and various ministries ’ websites.

9 Note that we explain changes in the rm's structure by the level of international

taxation. Expanding rms may have a need to establish new subsidiaries even if there

are no changes in the level of taxation.

10 The conditional logit model imposes the axiom of independence of irrelevant alter-

natives (IIA), which implies that adding a third option or changing the characteristics

of a third alternative does not affect the relative odds between any two options consid-

ered. As a robustness check, we have estimated a nested logit model that allows us to

relax the IIA assumption yielding results broadly similar to the ones reported in Table 5

(unreported).11 To enable estimation of the year xed effects, we simultaneously consider how a

multinational selects the location of its foreign subsidiaries over the entire period

1999–2003, rather than 1 year at a time.12 The possibility that multinationals' location choices might be correlated both geo-

graphically and over time has been suggested in a number of recent empirical studies.

Crozet et al. (2004), for instance, nd evidence of positive spillovers affecting the loca-

tion choices of French rms investing abroad suggesting that French multinationals

cluster geographically in a gradual process of agglomeration and market penetration.

Head and Mayer (2004) study the determinants of the location choices of Japanese

multinationals in Europe and nd a similar tendency for these multinational rms to

cluster in the same countries.


http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-


8/13

The corresponding estimate of the marginal effect of the effective tax

rate on the probability of location is−0.22, meaning thata one percent-

age point increase in theeffective taxrate (from a level of zero) reduces

the probability of location by 0.22%.

Next, regression 2 includes a host of non-tax control variables, and

it yields an estimated coef cient for the effective tax rate variable of

−4.29 that is signicant at the 1% level. The corresponding estimate

of the marginal effect of the effective tax rate on the probability of

location is −0.78. Among the controls, a country's relative GDP is es-

timated to increase the probability of subsidiary location with signif-

icance at the 1% level. Contiguous countries are also more likely to

receive foreign subsidiaries with signicance at the 1% level. The dif-

ference in labor costs variable enters with a negative coef cient that

is signicant at 1%, perhaps suggesting that high wages in prospectivehost countries signal sought after labor skills. The economic freedom

and EU membership variables both enter with positive coef cients

that are signicant at 10% and 1%, respectively.13

Regression 3 substitutes the host country corporate tax rate for the

effective tax rate. The estimated parameter on the host country tax var-

iablehas a value of −3.42and it is signicant at the 1%level. Inlinewith

this, a one percentage point increase in the host countrytax rate is esti-

mated to reduce the probability of location by 0.63%. The control vari-

ables enter regression 3 in qualitatively the same way as before.

Regression 4 in turn includes the international tax variable –

reecting both non-resident withholding taxation in the host country

and additional parent country corporate taxation – with an estimated

coef cient of −1.87 that is signicant at 5%. The corresponding mar-

ginal effect of the international tax rate on the probability of location

is estimated to be relatively small in absolute value at 0.19.

Next, regression 5 jointly includes the host country tax rate and

the international tax rate, yielding estimated coef cients of −

4.38and −3.93, that are both signicant at 1%, with corresponding esti-

mated marginal effects of −0.82 and −0.73.

Finally, regression 6 splits up the effective tax rate into its three com-

ponents: the host country corporate tax rate, the non-resident withhold-

ing tax rate, and the additional parent country corporate tax rate.

Parameter estimates for the host country tax rateand parent country cor-

porate tax rate are negative and statistically signicant at 1%, while the

non-resident dividend withholding tax rateobtains a negative coef cient

that is statistically insignicant. A one percentage point increase in the

host country and additional parent country tax rates are estimated to re-

duce theprobabilityof location by 0.90% and1.07% respectively,while the

analogous estimated effect of the non-resident withholding tax is 0.06%.

Our results suggest that host country and additional parent country

taxation both play a signi

cant role in multinationals' location choices.

13 The presence of intangible assets or R&D intensity could similarly affect the sensi-

tivity of international location decisions to taxation. Recent evidence provided by

Dischinger and Riedel (2011) shows that corporate taxation signicantly affects the in-

ternational location of intangible assets given a multinational rm's structure. To test

this, we distinguished between rms that belong to sectors with low and medium or

high R&D intensity using information from the OECD Science, Technology and Industry

Scoreboard (see OECD, 2003). Following the OECD taxonomy, medium or high R&D in-

tensity sectors have R&D expenditure exceeding 2% of total value added. We estimated

separate regressions analogous to regression 2 of Table 5 for the samples of rms be-

longing to the two groups of sectors, yielding very similar parameter estimates for

the effective tax rate variable (unreported).

Latvia Lithuania Luxembourg Malta Netherlands Norway Poland Portugal Romania Russia Slovak

Rep.

Slovenia Spain Sweden Switzerland Turkey United

Kingdom

0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

0 0 1 1 1 1 1 1 1 0 1 0 1 1 1 0 1

1 1 0 0 1 1 1 0 1 1 1 0 0 1 0 0 1

0 0 0 1 0 1 1 0 1 1 1 1 0 1 0 0 11 1 1 0 1 1 1 1 1 1 1 1 1 1 1 0 1

1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1

1 1 0 0 1 1 1 0 0 0 0 0 0 1 0 0 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

0 0 1 0 1 1 1 1 1 0 1 0 1 1 1 0 1

0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 1 1 0 1 1 1 1 0 0 0 0 1 1 1 0 1

1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 0 1

0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

X 1 0 0 1 1 1 0 1 0 1 1 0 1 1 0 1

1 X 0 0 1 1 1 0 1 0 1 1 0 1 1 0 1

0 0 X 1 1 1 1 1 1 1 1 1 1 1 1 0 1

1 0 1 X 1 1 1 1 1 0 1 0 0 1 1 0 1

1 1 1 1 X 1 1 1 1 1 1 1 1 1 1 1 1

1 1 1 1 1 X 1 1 1 1 1 1 1 1 1 1 1

1 1 1 1 1 1 X 1 1 1 1 1 1 1 1 1 10 0 1 1 1 1 1 X 1 1 1 0 1 0 1 0 1

1 1 1 0 1 1 1 1 X 1 1 1 1 1 1 1 1

0 0 1 0 1 1 1 1 1 X 1 0 1 1 1 1 1

1 1 1 1 1 1 1 1 1 1 X 1 1 1 1 1 1

1 1 1 0 1 1 1 0 1 1 1 X 1 1 1 0 1

0 0 1 0 1 1 1 1 1 1 1 1 X 1 1 0 1

1 1 1 1 1 1 1 1 1 1 1 1 1 X 1 1 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 X 0 1

0 1 0 0 1 1 1 0 1 1 1 0 0 1 0 X 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 X


http://-/?-http://-/?-http://-/?-http://-/?-


9/13

To evaluate the implications of tax rate changes in host and parent coun-

tries for location decisions,one needs to recognize thatsuch changes gen-

erally alter both our host country and additional parent country tax

variables. An increase in the host country corporate tax rate, for instance,

would lead to an increase in the host country tax variable and an

off-setting reduction in the additional parent country tax variable if a

full foreign tax credit is available (this is the case if the host country tax

rate remainsbelow theparent countrytax rate absent non‐resident with-

holding taxes). In this instance, the effective tax rate would remainunchanged, and the implied estimated impact on location would be

small, as the estimated marginal effects of the host and additional parent

country tax variables on location probabilities of −0.90 and −1.07 are

very similar. An increase in the host country corporate tax rate, however,

increases the host country tax variable while leaving the additional par-

ent country tax variable unchanged, if the host country tax rate exceeds

the parent country tax rate. In this instance, a higher host country corpo-

ratetax ratediscourages locationin the host country, while increasing the

combined, effective tax rate. Overall, our estimation is consistent with a

negative relationship between the host country corporate tax rate and

subsidiary location as generally reported in the literature surveyed by

De Mooij and Ederveen (2006).

Our nding of a large role for the additional parent country tax in

foreign subsidiary location is new, and perhaps surprising, as many

countries in the world allow parent country taxes on foreign source

income to be deferred until dividend repatriation, diminishing its po-

tential to affect location decisions. Similarly, some countries allow

so-called worldwide income averaging. This practice allows multina-

tionals resident in, for instance, the U.S. to claim foreign tax credits for

foreign taxes paid in high-tax countries against U.S. taxes due on in-

come from low-tax countries, potentially reducing the burden of par-

ent country corporate income taxation.

The result that foreign subsidiary location is sensitive to additional

parent country taxationreects thatparent country taxationis rather dis-

criminatory, as it only applies to parent rms residing in the pertinent

parent country.Additional parent country taxationthus discouragesmul-

tinational rm ownership of foreign subsidiaries that are subject to addi-

tional parent country taxation. The estimated negative impact of parent

country location on subsidiary location implies that multinationals acton an apparently signicant incentive to bring about an international

ownership pattern of subsidiaries that is internationally tax-ef cient.

Substantial withholding taxes also put particular foreign owners at a

comparative disadvantage at owning local assets vis-à-vis any other for-

eign owners that are lowly taxed and local owners. All the same, wend

that non‐resident dividend withholding taxes are statistically insigni-

cant in determining subsidiary location decisions. To explain this, rst

note that the EU Parent-Subsidiary Directive provides that no withhold-

ing tax shall be levied on dividend payments between related cross-

border companies. This makes the application of a withholding tax rare

in our dataset. Second, companies can potentially avoid withholding

taxes by creating conduit companies or by letting the subsidiary compa-

ny provide the parent company with a loan instead of a dividend.14

Next, Table 6 reports several robustness checks to gain additionalinsight in the impact of the international tax system on foreign

subsidiary location.15 First, we recognize that the calculation of our

effective tax variable is based on countries’ top statutory tax rates,

thus ignoring potential differences in tax base denitions regarding,

for instance, interest deductibility that equally affect the tax burden

in a particular country. As an alternative effective taxmeasure, regres-

sion 1 of Table 6 includes the Effective Average Tax Rate (EATR) for

outbound FDI as computed by ZEW (2008) starting from regression

2 of Table 5. The EATR reects statutory information on tax rates, in-

terest deductibility, as well as allowances for capital depreciation,

and it is constructed as the overall tax liability triggered by a projectdivided by the overall returns generated by the project in the absence

of taxation (see Devereux and Grif th, 1999, 2003). The EATR approx-

imatesan average taxon FDI, unlikeour effective tax rate based on top

statutorytax rates,and thus canbe expected to inuence international

location decisions. In regression 1, the EATR variable obtains a nega-

tive coef cient of −2.51 that is statistically insignicant. The marginal

effect in the EATR regression is −0.60. The ZEW (2008) data do not

provide a split up theEATR variable into host and parentcountry com-

ponents and hence we cannot perform regressions using EATR-like tax

variables analogous to regressions 3–6 of Table 5.

Next, we investigate whether the responsiveness of subsidiary lo-

cation is different for subsidiaries with a low or high return on assets.

Subsidiaries with a low return on assets presumably pay low taxes,

suggesting that there is a reduced role of taxation to affect loca-

tion choices. Alternatively, subsidiary location of foreign subsidiaries

with a high return on assets could be rather insensitive to taxation,

if the high return on assets reects the realization of rents that are lo-

cation and owner specic. To investigate this, columns 2 and 3 of

Table 6 report regressions for the samples of subsidiaries with a rate

of return on assets below and above the sample median. Otherwise,

these two regressions are analogous to regression 6 of Table 5. The

host country tax variable and the additional parent country tax

enter with negative and statistically signicant coef cients in both

regressions 2 and 3. Interestingly, the estimated marginal effects of

the host and additional parent country taxes on the probability of lo-

cation are both smaller in absolute terms for the high-ROA sample,

which suggests that subsidiaries with a high rate of return experience

rents that are location or owner specic.

Next, we consider whether the responsiveness of location is differ-ent for subsidiaries with low and high xed assets. Subsidiaries that

use high xed assets may be dif cult to relocate physically, and

hence one expects their location to be less sensitive to the host coun-

try tax. Regressions 4 and 5 are based on samples of subsidiaries with

ratios of xed assets to total assets below and above the median. In

regressions 4 and 5, the host country tax and the additional parent

country tax both enter with negative and statistically signicant coef-

cients (albeit at different levels of signicance). The implied margin-

al effects of both tax rates on the probability of location are relatively

large in absolute terms for the sample of subsidiaries with low xed

assets. These results suggest that the location of subsidiaries with

low xed assets is relatively more sensitive to host and parent coun-

try taxation.16

As discussed, the option to defer parent country taxes may makeforeign subsidiary location less responsive to the additional parent

country tax. Previously, Huizinga and Voget (2009) have collected in-

formation on whether the deferral option is available for parent rms

located in a particular home country. We can use this information to

check whether the deferral option indeed mutes the responsiveness

of location to additional parent country taxation. To start, regression

6 in Table 6 introduces an interaction of the effective tax and a

14 The conversion of dividend payments into a loan to the parent potentially also

eliminates parent country taxation on repatriated income. Under US CFC rules, howev-

er, such loans could be labelled ‘deemed dividends’ and trigger parent country taxa-

tion. In the EU, transactions of this kind may be subject to transfer pricing rules, but

adjustments of the transfers to qualify as dividends and subsequent adjustments of

parent country taxation seem to be the exception rather than the rule.15 As additional robustness checks, we have performed regressions that inter alia (i)

exclude holding companies, (ii) include intermediate companies that are both parent

and subsidiary, and (iii) include tax variable interactions with a dummy variable sig-

naling a previously established subsidiary in the country. Our main results on the im-

pact of the international taxsystem on foreign subsidiary location are unaltered in each

of these regressions (unreported).

16 We tested for the equality of coef cients on the tax variable across the two regres-

sions that result from dividing the sample into low ROA/high ROA and low xed assets/

high xed assets subsamples as reported in Table 6 using a Wald test. The results indi-

cate that only the withholding tax variable obtains signicantly different coef cients in

each of the two pairs of regressions. In all four regressions, the withholding tax variable

tax variable, however, obtains a coef cient that itself is insignicant.


http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-


10/13

deferral option dummy in regression 2 of Table 5. The effective tax

variable now obtains a coef cient of −9.17 that is signicant at 1%

and lower than the corresponding coef cient of −

4.29 in regression

2 of Table 5. The interaction of the effective tax with the deferral op-

tion dummy obtains a positive coef cient of 4.85 that is signicant at

1%. The corresponding marginal effects on the probability of location

Table 4

Descriptive statistics for subsidiaries of European multinationals.

Panel A. Number of parent companies and subsidiaries used in basic regression

Country Number of parent companies by home country Number of subsidiaries

By home country By host country

Austria 9 9 6

Belgium 29 36 34

Bulgaria 9 11 2

Croatia 0 0 1Cyprus 0 0 0

Czech Republic 0 0 5

Denmark 26 28 110

Estonia 5 6 1

Finland 19 23 30

France 84 116 36

Germany 56 75 97

Greece 33 48 4

Hungary 0 0 6

Iceland 2 3 4

Ireland 9 11 39

Italy 64 78 38

Latvia 1 2 2

Lithuania 0 0 0

Luxembourg 1 1 2

Netherlands 64 89 51

Norway 33 45 74Poland 5 6 39

Portugal 19 21 11

Romania 0 0 9

Russia 1 1 0

Slovak Republic 1 1 0

Slovenia 6 6 0

Spain 49 59 88

Sweden 69 76 68

Switzerland 13 14 6

Turkey 0 0 0

United Kingdom 115 144 146

Total 722 909 909

Panel B. Summary statistics for variables in basic regression

Variable n Mean Standard dev. Min Max

Subsidiary location 26,648 0.034 0.181 0 1

Effective tax 26,648 0.353 0.073 0.125 0.750International tax 26,648 0.051 0.070 0 0.550

Host country tax 26,648 0.302 0.083 0 0.567

GDP bilateral 26,648 0.033 0.054 0.0003 0.264

Contiguity 26,648 0.166 0.371 0 1

Difference in labor costs 26,648 0.270 0.563 −2.098 2.373

Economic freedom 26,648 6.430 1.021 3.800 8.425

EU membership 26,648 0.464 0.499 0 1

Panel C. Correlation matrix for variables in basic regression

Subsidiary

location

Effective

tax

International

tax

Host

country tax

GDP

bilateral

Contiguity Difference in

labor costs

Economic

freedom

EU

membership

Sub sidiar y loca tion 1 .0 00 0

Effective tax −0.0059 1.0000

International tax −0.0798** 0.3369** 1.0000

Host country tax 0.0628** 0.5976** −0.5536** 1.0000

GDP bilateral 0.1669** 0.2662** −

0.2907** 0.4830** 1.0000Contiguity 0.0733** 0.0794** −0.1098** 0.1638** 0.2324** 1.0000

Difference in labor costs −0.0923* 0.0397* 0.1522* −0.0945* −0.2416 −0.1455 1.0000

Economic f reedom 0 .1 03 0** −0.1442** −0.1667** 0.0144** 0.1943** 0.0794** −0.6670* 1.0000

EU membership 0.1402** 0.0622** −0.4990** 0.4800** 0.4975** 0.2092** −0. 4547** −0.4891* 1.0000

Notes: Subsidiary location is a dummy variable equaling 1 if a subsidiary is located in a country and 0 otherwise. Effective tax is the tax rate on dividend income generated in the potential

subsidiary country resulting from corporate income taxation in host and parent countries and non-resident withholding taxation in the host country. International tax is the difference be-

tweenthe effectivetax andthe host country corporatetax.Host country taxis thecorporate incometax in thesubsidiary country, includinglocaltaxes andpossiblesurcharges.GDP bilateral

is the ratio of the GDP in a potential host country to the sum of GDP of all potential host countries. Contiguity is a binary variable equal to 1 if the parent and potential host countries have a

commonborder. Difference in labor costs is thelog of theratioof labor costs in thehome country andlaborcosts in thepotential host country. Economic freedom is an index scaledbetween

0and10reectingthe followingFraser indicators of thepotential host country: soundnessof legalsystem, absence of trade barriers,and absence of price controls.EU membership is a binary

variable equal to 1 if the potential host country is a member of the European Union. Basic regression refers to regression 2 in Table 5. * denotes signicance at 5%; ** signicance at 1%.



11/13

are −1.63 and 0.86.These resultssuggestthat the deferral option re-

duces the responsiveness of location to the effective country tax by

about half.

In analogous fashion, regression 7 of Table 6 includes an interac-

tion of the international tax variable with the deferral option

dummy in regression 5 of Table 5. The international tax variable

and its interaction with the deferral option dummy enter with neg-ative and positive coef cients, respectively, that are signicant at

1% and 5% respectively. The implied marginal effects of the two var-

iables on the probability of location are −3.10 and 2.48, which sug-

gests that the deferral option reduces the responsiveness of location

to the international tax variable by 80%. Finally, regression 8 of

Table 6 includes an interaction of the additional parent country

tax with the deferral option dummy in regression 6 of Table 5.

The additional parent country tax obtains a negative coef cient

that is signicant at 5%, while its interaction with the deferral op-

tion dummy obtains a positive coef cient that is statistically insig-

nicant. Estimated marginal effects of the two variables on the

probability of location suggest that the deferral option reduces the

responsiveness of location to the additional parent country tax by

about 72%.

Overall, the robustness checks conrm that international double

taxation, as reected in our international tax and additional parent

country tax variables, is important in explaining location decisions re-

garding new foreign subsidiaries. More specically, our results sug-

gest that foreign subsidiary location is relatively sensitive to

international double taxation for subsidiaries with a low return on as-

sets, low xed assets and in the absence of the deferral option.

5. Conclusions

This paper provides evidence on the implications of international

taxation for the location of the foreign subsidiaries of multinational

rms. Our analysis uses panel data on the structure of multinational

rms in 33 European countries over the period 1999–2003. This rich

data set allows us to estimate the separate effects of host country

corporate income taxation, host country non-resident dividend with-

holding taxation, and additional parent country corporate income

taxation on the location decisions of multinational rms.

Our main result is that the additional parent country corporate

taxation of foreign source income has an independent, strongly nega-

tive effect on the probability of foreign subsidiary location in potential

Table 5

Taxation and foreign subsidiary location. The dependent variable is subsidiary location equaling 1 if a subsidiary is located in a country and 0 otherwise. Effective tax is the tax rate

on dividend income generated in the potential subsidiary country resulting from corporate income taxation in host and parent countries and non-resident withholding taxation in

the host country. Host country corporate tax is the corporate income tax in the subsidiary country, including local taxes and possible surcharges. International tax is the difference

between the effective tax and the host country corporate tax. Withholding tax is the non-resident withholding tax burden imposed by the host country. Additional parent country

corporate tax is the additional corporate tax burden imposed by parent country after double tax relief. GDP bilateral is the ratio of the GDP in a potential host country to the sum of

GDP of all potential host countries. Contiguity is a binary variable equal to 1 if the parent and potential host countries have a common border. Difference in labor costs is the log of

the ratio of labor costs in the home country and labor costs in the potential host country. Economic freedom is an index scaled between 0 and 10 reecting the following Fraser

indicators of the potential host country: soundness of legal system, absence of trade barriers, absence of price controls, and transfers and subsidies as a share of GDP. EU member-

ship is a binary variable equal to 1 if the potential host country is a member of the European Union. All regressions include year xed effects that are not reported. Sample reects

location decisions of new foreign subsidiaries. Estimation is by conditional logit model. Standard errors that are heteroskedasticity-consistent and adjusted for clustering at the levelof the host country are reported between brackets. Marginal effect is the slope of the probability curve with respect to an included tax variable evaluated at zero while other ex-

planatory variables are evaluated at their mean. * denotes signicance at 10%; ** signicance at 5% and *** signicance at 1%.

(1) (2) (3) (4) (5) (6)

Variables (expected sign) Tax

only

Basic

regression

Host country

corporate tax

International

taxation

Host country corporate and

international taxes

Withholding and parent

country corporate tax

Effective tax (−) −0.871 −4.292***

( 0.57 4) ( 0.63 4)

Host country corporate tax (−) −3.424*** −4.376*** −4.806***

(0.500) (0.624) (0.654)

International tax (−) −1.869** −3.927***

(0.938) (1.042)

Withholding tax (−) −0.322

(1.400)

Additional parent country corporate

tax (−)

−5.731***

(1.400)

GDP bilateral (+) 7.420*** 7.196*** 5.404*** 7.470*** 7.620***(0.575) (0.548) (0.475) (0.566) (0.571)

Contiguity (+) 0.396*** 0.372*** 0.372*** 0.395*** 0.402***

(0.084) (0.085) (0.086) (0.084) (0.084)

Differences in labor costs (+/−) −0.692*** −0.768*** −0.573*** −0.705*** −0.737***

(0.179) (0.190) (0.174) (0.186) (0.181)

Economic freedom (+) 0.140* 0.139* 0.291*** 0.134* 0.133*

(0.073) (0.073) (0.067) (0.074) (0.074)

EU membership (+) 0.952*** 1.192*** 0.814*** 0.981*** 1.107***

(0.105) (0.119) (0.117) (0.123) (0.129)

Observations 26,648 26,648 26,648 26,648 26,648 26,648

Pseudo R-squared 0.001 0.135 0.133 0.127 0.136 0.137

Marginal effects of tax variables

Effective tax −0.218 −0.784

Host country corporate tax −0.632 −0.816 −0.895

International tax −0.189 −0.732

Withholding tax −

0.060Additional parent country

corporate tax

−1.070



12/13

host countries, despite the fact that parent country taxation can gen-

erally be deferred until income is repatriated. This result may reect

that a parent country's taxation is rather discriminatory as it only ap-

plies to parent rms residing in the parent country. The high sensitiv-

ity of foreign subsidiary location to parent country taxation suggests

that this tax is particularly distortive. Paradoxically, the parent coun-

try tax may be highly distortive on account of the foreign credit

mechanism – aiming to alleviate international double taxation – as

the foreign tax credit mechanism produces a high variability of the

(post credit) parent country tax across foreign location choices. The

sensitivity of foreign subsidiary location to parent country taxation

strengthens the case for abolishing worldwide taxation in favor of

territorial taxation.

Additional parent country taxation may not only distort the for-

eign subsidiary location decision, but also the capital investment de-

cision once location is determined. The impact of parent country

taxation on capital investment abroad would be an interesting area

for future research.

Table 6

Taxation and foreign subsidiary location: robustness tests. The dependent variable is subsidiary location equaling 1 if a subsidiary is located in a country and 0 otherwise. Effective

tax is the tax rate on dividend income generated in the potential subsidiary country resulting from corporate income taxation in host and parent countries and non-resident with-

holding taxation in the host country. Host country corporate tax is the corporate income tax in the subsidiary country, including local taxes and possible surcharges. International

tax is the difference between the effective tax and the host country corporate tax. Withholding tax is the non-resident withholding tax burden imposed by the host country. Ad-

ditional parent country corporate tax is the additional corporate tax burden imposed by parent country after double tax relief. GDP bilateral is the ratio of the GDP in a potential host

country to the sum of GDP of all potential host countries. Contiguity is a binary variable equal to 1 if the parent and potential host countries have a common border. Difference in

labor costs is the log of the ratio of labor costs in the home country and labor costs in the potential host country. Economic freedom is an index scaled between 0 and 10 re ecting

the following Fraser indicators of the potential host country: soundness of legal system, absence of trade barriers, absence of price controls, and transfers and subsidies as a share of

GDP. EU membership is a binary variable equal to 1 if the potential host country is a member of the European Union. All regressions include year xed effects that are not reported.

The EATR is the cross-border Average Effective Corporate Tax Rate taken from ZEW (2008). Deferral is a dummy variable equal to 1 if deferral is possible and zero otherwise. Re-gressions 2 and 3 split the sample into subsidiaries that have respectively a ratio of pre-tax prots to total assets below or equal to and above the median value. Regressions 4 and 5

split the sample into subsidiaries that have respectively a ratio of xed assets to total assets below or equal to and above the median value. Sample re ects location decisions of new

foreign subsidiaries. Standard errors thatare heteroskedasticity-consistent and adjusted for clustering at the level of the host country are reported between brackets. Marginal effect

is the slope of the probability curve with respect to an included tax variable evaluated at zero while other explanatory variables are evaluated at their mean. * denotes signicance at

10%; ** signicance at 5% and *** signicance at 1%.

(1) (2) (3) (4) (5) (6) (7) (8)

Variables (expected sign) EATR Low ROA High ROA Low xed

assets

High xed

assets

Deferral Deferral

two-way

split

Deferral

three-way

split

EATR −2.513

(1.883)

Effective tax (−) −9.166***

(1.798)

Host country corporate

tax (−)

−4.367*** −5.238*** −5.911*** −3.930*** −4.504*** −4.882***

(0.806) (0.912) (0.955) (0.869) (0.622) (0.652)

International tax (−

) −

16.61***(5.620)

Withholding tax (−) 0.383 −1.330 1.036 −1.114 −0.202

(1.725) (2.068) (2.191) (1.627) (1.400)

Additional parent country

corporate tax (−)

−5.951***

(1.802)

−5.594***

(1.982)

−6.803**

(3.239)

−5.277***

(1.475)

−17.850**

(8.167)

Deferral× Effective tax (+) 4.853***

(1.711)

Deferral×International

tax (+)

13.26**

(5.580)

Deferral×Additional parent

country tax (+)

12.86

(8.180)

GDP bilateral (+) 4.765*** 7.066*** 8.094*** 9.367*** 5.980*** 7.365*** 7.478*** 7.616***

(0.810) (0.779) (0.772) (0.811) (0.771) (0.574) (0.561) (0.568)

Contiguity (+) 0.664*** 0.448*** 0.347*** 0.408*** 0.396*** 0.412*** 0.411*** 0.416***

(0.109) (0.114) (0.119) (0.122) (0.113) (0.084) (0.084) (0.084)

Labor cost (+/−) −0.197 −1.027*** −0.353 −1.081*** −0.575*** −0.682*** −0.713*** −0.731***

(0.693) (0.199) (0.300) (0.330) (0.214) (0.178) (0.186) (0.180)Eco nomic free dom (+) 0.038 −0.041 0.351*** 0.048 0.183* 0.149** 0.133* 0.133*

(0.280) (0.080) (0.104) (0.113) (0.094) (0.072) (0.074) (0.073)

EU membership (+) 1.446*** 0.902*** 1.407*** 1.552*** 0.849*** 0.897*** 0.977*** 1.112***

(0.159) (0.168) (0.185) (0.213) (0.158) (0.107) (0.126) (0.131)

Observations 12,175 14,176 12,384 11,658 14,990 26,648 26,648 26,648

Pseudo R-squared 0.108 0.119 0.170 0.192 0.106 0.137 0.137 0.138

Marginal effects of tax variables

Effective tax −0.603 −1.625

Host country corporate tax −1.093 −0.366 −1.342 −0.640 −0.841 −0.910

International tax −3.100

Withholding tax 0.097 −0.094 0.237 −0.174 −0.037

Additional parent country

corporate tax

−1.486 −0.391 −1.545 −0.854 −3.329

Deferral×Included

tax variable

0.860 2.475 2.398



13/13

Appendix A

References

Altshuler, R., Grubert, H., 2001. Where will they go if we go territorial? Dividend ex-emption and the location decisions of U.S. multinational corporations. NationalTax Journal 54, 787–809.

Barrios, S., Görg, H., Strobl, E., 2005. Foreign direct investment, competition and indus-trial development in the host country. European Economic Review 49, 1761 –1784.

Buettner, T., Ruf, M., 2007. Tax incentives and the location of FDI: evidence froma panelof German multinationals. International Tax and Public Finance 14, 151–164.

Crozet, M., Mayer, T., Mucchielli, J.-L., 2004. How do rms agglomerate? A study of FDIin France. Regional Science and Urban Economics 34, 27–54.

De Mooij, R., Ederveen, S., 2006. What a difference does it make? Understanding theempirical literature on taxation and international capital ows. Economic PapersNo. 261, European Commission. D-G for Economic and Financial Affairs, Brussels.

Desai, M., Foley, C., Hines, J., 2001. Repatriation taxes and dividend distortions. NationalTax Journal 54, 829–851.

Desai, M.A., Hines, J.R., 2002. Expectations and expatriations: tracing the causes andconsequences of corporate inversions. National Tax Journal 55, 409–440.

Devereux, M.P., Grif th, R., 1998. Taxes and the location of production: evidence from apanel of US multinationals. Journal of Public Economics 68, 335–367.

Devereux, M.P., Grif th, R., 1999. The taxation of discrete investment choices. WorkingPaper No. W98/16. Institute for Fiscal Studies, London.

Devereux, M.P., Grif th, R., 2003. Evaluating tax policy for location decisions. Interna-tional Tax and Public Finance 10, 107–126.

Devereux, M.P., Maf ni, G.,2007. The impact of taxation on the location of capital,rmsand prot: a survey of empirical evidence. Oxford University Centre for Business

Taxation Working Paper 07/02.Dischinger, M., Riedel, N., 2011. Corporate taxes and the location of intangible assets

within multinational rms. Journal of Public Economics 95, 691–707.Egger, P., Loretz, S., Pfaffermayr, M., Winner, H., 2009. Bilateral effective tax rates and

foreign direct investment. International Tax and Public Finance 16, 822 –849.Görg, H., Greenaway, D., 2004. Much ado about nothing? Do domestic rms really ben-

et from foreign direct investment? World Bank Research Observer 19, 171 –197.

Hartman, D.G., 1985. Tax policy and foreign direct investment. Journal of Public Eco-nomics 26, 107–121.

Head, K., Mayer, T., 2004. Market potential and the location of Japanese rms in theEuropean Union. Review of Economics and Statistics 86, 959–972.

Hines, J.R., 1996. Altered states: taxes and the location of foreign direct investment inAmerica. American Economic Review 86, 1076–1094.

Huizinga, H., Voget, J., 2009. International taxation and the direction and volume of cross-border M&As. Journal of Finance 64, 1217–1249.

Huizinga, H., Laeven, L., Nicodème, G., 2008. Capital structure and international debtshifting. Journal of Financial Economi

Date post:	05-Jul-2018
Category:	Documents
Upload:	carlos-martins
View:	216 times
Download:	0 times

Barrios Et Al. (2012) International Taxation and Multinational Firm Location Decisions

Documents